Difficulties with memory are among the most common complaints in the veteran population. The proposed work involves three related studies of the organization and neurology of human memory systems. The declarative memory system supports the capacity to learn and remember facts and events and depends on the hippocampus. Historically, declarative memory has been thought to represent conscious memory, i.e., knowledge that is accessible to awareness. Indeed, accessibility to awareness has been thought to be a defining feature of declarative memory. However, the characteristics of declarative (and nondeclarative) memory remain active topics of research, and the link between declarative memory, hippocampal function, and conscious knowledge is uncertain. One recent proposal is that memory can be hippocampus-dependent but can at the same time be inaccessible to conscious awareness. These issues became especially prominent following the discovery that eye movements can reflect a form of memory. For example, individuals scan a scene differently depending on whether the scene is novel or familiar. They make fewer fixations, sample fewer regions, and also (from our Preliminary Studies) sample different regions when viewing a familiar (rather than a novel) scene. What kind of memory effects are these? Do these eye movement effects simply provide another measure of conscious, declarative memory? Alternatively, are they automatic and independent of awareness that a scene is novel or familiar? Furthermore, are they hippocampus-dependent or hippocampus-independent? The nature of eye movements has potential clinical relevance for veterans with limited response capability. These questions will be addressed in three experiments. The first experiment (1a) aims to determine whether experience-dependent eye movements vary with the strength of declarative memory (four levels of strength will be established by varying the number of scenes to be remembered and the retention interval across four groups of participants). Experiment 1b will test well-characterized memory-impaired patients with bilateral hippocampal lesions to ask whether these experience-dependent eye movements are hippocampus-dependent. The second experiment (2a) introduces a novel experimental design to ask, scene by scene, whether experience-dependent eye movements occur only when individuals consciously (explicitly) recognize a scene as familiar, or whether these eye movements are automatic and occur independently of the conscious experience of recognition. Experiment 2b will test memory-impaired patients to ask, under different conditions than Experiment 1b, whether experience dependent eye movements are hippocampus-dependent. The third experiment turns to functional magnetic resonance imaging (fMRI), using an improved eye movement tracker compatible with the fMRI environment. Hippocampal activity increases when familiar scenes are presented and is even greater when scenes are recognized with high confidence. Here the question is whether hippocampal activity, recorded in response to familiar scenes, correlates with the eye movement effects. Is there a tight relationship between these two measures, or do eye movements to familiar vs. novel scenes occur independently of hippocampal activity. Through these three experiments, it should be possible to decide what kind of memory is expressed through eye movements, whether it is conscious or unconscious memory, and whether it is hippocampus-dependent or independent. These questions are foundational to understanding the memory systems of the human brain.